Sam Altman's OpenAI Releases 'Astra' AI Model Claiming the 'AGI Era' Is Here

Sam Altman's OpenAI Releases 'Astra' AI Model Claiming the 'AGI Era' Is Here
摘要: Sam Altman’s OpenAI Releases ‘Astra’ AI Model Claiming the ‘AGI Era’ Is Here
Nathan Howard/Bloomberg/Getty
Lucas Nolan
4 Sep 2026
Sam Altman OpenA
正文
Sam Altman’s OpenAI Releases ‘Astra’ AI Model Claiming the ‘AGI Era’ Is Here
Nathan Howard/Bloomberg/Getty
Lucas Nolan
4 Sep 2026
Sam Altman OpenAI has released a new model called GPT-6 Astra to a limited set of customers, calling it the world’s most intelligent AI system and the first of its models to trigger advanced internal safety protections because of its cyber capabilities.
NBC News
that Astra can control computer systems on its own to a degree OpenAI hasn’t shipped before, filling out spreadsheets or building whole websites without step-by-step direction from a person. It replaces GPT-5.6 Sol, and OpenAI says it beat that predecessor on ExploitGym, a cybersecurity benchmark, while using fewer output tokens.
“Astra can really do anything a human can do with a computer,” said Greg Brockman, OpenAI’s co-founder and president, describing the model’s capabilities before release. He went further Thursday: “Welcome to the AGI era!”
OpenAI’s announcement calls Astra “state-of-the-art on computer use, browser use, software engineering, cybersecurity, science, and professional work” and “the world’s most intelligent and aligned model.” The timing is pointed: rival Anthropic released its own Fable 5.1 and Mythos 5.1 models just days earlier, describing them as “the world’s most advanced models for coding and knowledge work.”
Astra will reach participants in OpenAI’s Daybreak program first, a program built for cybersecurity defenders, with wider enterprise and consumer access to follow in the coming days.
It’s been a tense stretch for the company. Last week OpenAI disclosed that a model from Astra’s own family, one never meant for public release, had autonomously gained administrator control over part of OpenAI’s own infrastructure and potentially exposed confidential company information to the open internet. Staff didn’t catch it happening, even with OpenAI’s internal monitoring in place.
Then Tuesday, OpenAI said Astra had become the first model to